{
  "docId": "019dd923-5e88-73ef-bd5d-0856e1444fb9",
  "docSlug": "8757f1b44ef7f176",
  "documentTitle": "2024 Air Street Capital The State of AI Report 2024",
  "authorId": "AirStreetCapital",
  "authorName": "Air Street Capital",
  "documentKindSlug": "consulting-deck",
  "documentKindLabel": "Consulting deck",
  "sourceTypeSlug": "vc_research",
  "sourceTypeLabel": "VC research",
  "presentationDate": null,
  "orientation": "landscape",
  "aspectRatio": 1.777,
  "pageNumber": 202,
  "pageCount": 213,
  "prevPage": 201,
  "nextPage": 203,
  "slideType": "industry_trends",
  "function": "summarize",
  "density": "dense",
  "nDataPoints": 15,
  "notes": "The slide highlights the ongoing debate and lack of consensus on whether LLMs significantly increase biological threat risks compared to standard internet access.",
  "elementsJson": [
    "headline_text",
    "bullet_list",
    "bar_chart_grouped"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5d-0856e1444fb9/202",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9#slide-202",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "While classifying o1 as a 'medium' biorisk (a first for an OpenAI model), the company said that \"the model cannot yet automate biological agentic tasks\".",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-47c3-7407-8d22-63e0f7db0c55",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.45,
        "w": 0.4,
        "x": 0.55,
        "y": 0.42
      },
      "kind": "chart",
      "text": "Biothreat Information Long-Form performance comparison across GPT-4o, o1-mini, and o1-preview models.",
      "attrs": null,
      "subkind": "bar-grouped",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "4a89f8d2-5095-4c3e-8579-043dd7c937b3",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.65,
        "w": 0.48,
        "x": 0.05,
        "y": 0.22
      },
      "kind": "list",
      "text": "Anthropic captured headlines in the summer of 2023, after an unpublished study highlighted the potential of LLMs to accelerate the misuse of biology relative to internet access. Other labs have struggled to replicate this.\nOpenAI assessed uplifts in performance on a 10-point scale in biological threat creation with GPT-4 access versus an internet-only baseline. They found experts scored a 0.82 uplift and students a 0.41.\nWhile classifying o1 as a 'medium' biorisk (a first for an OpenAI model), the company said that \"the model cannot yet automate biological agentic tasks\". While it performed significantly better than 4o on biothreat information questions, it performed poorly on actual ideation.\nA RAND Corporation study concluded that current LLMs did not meaningfully change the operational risk of a biological weapons attack versus standard Internet access.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "572b1ddd-b5b7-49ee-8449-f1577b634f04",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": null,
      "kind": "metric",
      "text": "Accuracy (avg of 200 trials)",
      "attrs": null,
      "subkind": "primary",
      "toolName": "Quantification",
      "toolSlug": "quantification",
      "confidence": null,
      "componentId": "019dd952-47c3-7407-8d22-665a62a8155a",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.05,
        "w": 0.5,
        "x": 0.05,
        "y": 0.14
      },
      "kind": "title",
      "text": "Despite a storm of controversy, LLM biorisk remains uncertain...",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "c4b0b291-b9f5-4c11-abd9-989a20324c14",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.03,
        "w": 0.6,
        "x": 0.05,
        "y": 0.03
      },
      "kind": "title",
      "text": "Introduction | Research | Industry | Politics | Safety | Predictions",
      "attrs": null,
      "subkind": "subtitle",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "0ab5a978-c25d-4c5c-9170-8e70fabcd7d5",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [
    {
      "name": "Action Titles",
      "slug": "action-titles",
      "agent": "Architect",
      "layer": "slide",
      "matchId": "019dd95a-1055-74e0-a95a-deaa4f2749c0",
      "evidence": "Title 'Despite a storm of controversy, LLM biorisk remains uncertain...'",
      "confidence": 80
    }
  ],
  "frameworks": [],
  "arcBeats": [
    {
      "to": 204,
      "from": 154,
      "beatId": "019dd95a-0682-776c-8e35-605159a069e6",
      "arcName": "The Triple Take",
      "arcSlug": "triple-take",
      "beatName": "The Implications (So What)",
      "beatSlug": "triple-take-the-implications-so-what",
      "evidence": "Politics + Safety sections explore consequences and risks",
      "position": 2,
      "confidence": 78,
      "parentBeatName": "Reflection",
      "parentBeatSlug": "reflection"
    },
    {
      "to": 204,
      "from": 175,
      "beatId": "019dd95a-0682-776c-8e35-74f05e198faf",
      "arcName": "Voyage and Return",
      "arcSlug": "voyage-return",
      "beatName": "The Return",
      "beatSlug": "voyage-return-the-return",
      "evidence": "Safety section returns to risks/governance debate",
      "position": 4,
      "confidence": 55,
      "parentBeatName": "Resolution",
      "parentBeatSlug": "resolution"
    }
  ],
  "loops": [
    {
      "to": 204,
      "from": 202,
      "name": "Pattern Hunter",
      "slug": "02-pattern-hunter",
      "bestFor": "Time-pressed audiences, building consensus, when data is strong",
      "matchId": "019dd95a-07fe-70ce-8d3e-7d534fdd9d66",
      "evidence": "o1 medium biorisk, specialist tool risks, GenAI misuse zoom-out",
      "position": 24,
      "objective": "Refocus safety conversation on biorisk and everyday misuse",
      "structure": "Evidence A -> Evidence B -> Evidence C -> Pattern/Conclusion",
      "confidence": 65,
      "description": "Group multiple pieces of evidence that together point to a pattern or conclusion"
    }
  ],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}